Brand analysis using interactions with search result items

• US 8,639,559 B2
• Filed: 04/09/2012
• Issued: 01/28/2014
• Est. Priority Date: 04/09/2012
• Status: Active Grant

• Current Assignee: DoorDash, Inc.
• Original Assignee: International Business Machines Corporation
• Inventors: Kandogan, Eser, Mahmud, Jalal U., Smith, Barton A.
• Primary Examiner(s): Miller, Alan S
• Application Number:
  US13/442,348
• Publication Number:
  US 20130268314A1
• Time in Patent Office: 659 Days
• Field of Search: 705 711- 742, 707748-781
• US Class Current: 705/7.29
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

Embodiments of the invention relate to performing brand analysis based on user'"'"'s interactions with search results. In one embodiment, at least one search result list is analyzed. The search result includes a set of items corresponding to at least one search request submitted by at least one user. The set of items in the at least one search result list is determined to include at least one item corresponding to a brand analysis target item, and at least one other item. A selection by the at least one user of one of the brand analysis target item and the other item from the at least one search result list is recorded for the at least one search result list. A brand analysis metric is generated based on the recording. The brand analysis metric indicates a ratio of selections between the brand analysis target item and the other item.

Claims

1. A system comprising:

memory;

a processor communicatively coupled to the memory; and

a brand analysis manager communicatively coupled to the memory and the processor, the brand analysis manager is configured to;

analyze a set of web browsing interaction information;

determine, based on the analyzing, that at least one search process was performed by at least one user;

access, based on the set of web browsing interaction information, a web page comprising a search result list associated with the at least one search process, the search result list comprising a set of items corresponding to at least one search process;

analyze the at least one search result list;

determine that the set of items in the at least one search result list comprises at least one item corresponding to a brand analysis target item and also comprises at least one other item;

determine, based on the web browsing interaction information, that the at least one user performed a selection of one of the brand analysis target item and the other item from the at least one search result list;

record the selection by the at least one user; and

generate a brand analysis metric based on the recording, wherein the brand analysis metric indicates a ratio of selections by the at least one user between the brand analysis target item and the at least one other item.

2. The system of claim 1, wherein the brand analysis manager is configured to determine that the at least one user performed the selection based on:

analyzing the set of web browsing interaction information;

identifying, based on the analyzing, at least one action performed by the at least one user with respect to the at least one search result list; and

determining that the at least one action indicates the selection by the at least one user of one of the brand analysis target item and the other item from the search results list.

3. The system of claim 1, wherein the brand analysis manager is further configured to:

identify the at least one search result list in the web page based on;

analyzing the set of web browsing interaction information, wherein the set of web browsing interaction information comprises a plurality of actions performed by the at least one user with respect to web pages in a website;

identifying, based on the analyzing, a search action performed by the at least one user, wherein the search action corresponds to the at least one search process; and

retrieving a web page associated with the action, as indicated by the set of web browsing interaction information, performed by the at least one user immediately after the search action, wherein the web page is the web page comprising the at least one search list.

4. The system of claim 3, wherein identifying the at least one search result list in the web page is further based on:

determining a presentation pattern of search results for the web page;

constructing a document object model (DOM) tree of web page;

analyzing the DOM tree with respect to the presentation pattern; and

identifying, based on the analyzing, a DOM node in the DOM tree comprising the presentation pattern based on a similarity between child nodes of the DOM node, wherein the DOM node corresponds to the at least one search result list, and wherein each child node of the DOM node corresponds to an item in the at least one search result list.

5. The system of claim 1, wherein the brand analysis manager is configured to record by:

increasing a counter associated with the one of the brand analysis target item and the other item from the search results list in response to the selection thereof by the user.

6. The system of claim 1, wherein the brand analysis metric indicates a ratio between selections of the brand analysis target item and the other item when the brand analysis target item and the other item were presented in the same search result list.

7. A non-transitory computer program product comprising a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code configured to:

analyze a set of web browsing interaction information;

determine, based on the analyzing, that at least one search process was performed by at least one user;

access, based on the set of web browsing interaction information, a web page comprising a search result list associated with the at least one search process, the search result list comprising a set of items corresponding to at least one search process;

analyze the at least one search result list;

determine that the set of items in the at least one search result list comprises at least one item corresponding to a brand analysis target item and also comprises at least one other item;

determine, based on the web browsing interaction information, that the at least one user performed a selection of one of the brand analysis target item and the other item from the at least one search result list;

record the selection by the at least one; and

generate a brand analysis metric based on the recording, wherein the brand analysis metric indicates a ratio of selections by the at least one user between the brand analysis target item and the at least one other item.

8. The non-transitory computer program product of claim 7, wherein the computer readable program code is configured to determined that the at least one user performed the selection based on:

analyzing the set of web browsing interaction information,

identifying, based on the analyzing, at least one action performed by the at least one user with respect to the at least one search result list; and

determining that the at least one action indicates the selection by the at least one user of one of the brand analysis target item and the other item from the search results list.

9. The non-transitory computer program product of claim 7, wherein the computer readable program code is further configured to:

identify the at least one search result list in the web page based on;

analyzing the set of web browsing interaction information, wherein the set of web browsing interaction information comprises a plurality of actions performed by the at least one user with respect to web pages in a website;

identifying, based on the analyzing, a search action performed by the at least one user, wherein the search action corresponds to the at least one search process; and

retrieving a web page associated with the action, as indicated by the set of web browsing interaction information, performed by the at least one user immediately after the search action, wherein the web page is the web page comprising the at least one search list.

10. The non-transitory computer program product of claim 9, wherein the computer readable program code is further configured to identify the at least one search result list in the web page based on:

determining a presentation pattern of search results for the web page;

constructing a document object model (DOM) tree of web page;

analyzing the DOM tree with respect to the presentation pattern; and

identifying, based on the analyzing, a DOM node in the DOM tree comprising the presentation pattern based on a similarity between child nodes of the DOM node, wherein the DOM node corresponds to the at least one search result list, and wherein each child node of the DOM node corresponds to an item in the at least one search result list.

11. The non-transitory computer program product of claim 7, wherein the computer readable program code is configured to record by:

increasing a counter associated with the one of the brand analysis target item and the other item from the search results list in response to the selection thereof by the user.

12. The non-transitory computer program product of claim 7, wherein the brand analysis metric indicates a ratio between selections of the brand analysis target item and the other item when the brand analysis target item and the other item were presented in the same search result list.

13. The non-transitory computer program product of claim 7, wherein the other item is a competitor item of the target item.